Comparison · Data APIs

CoachLayer vs Terra and Sahha: three layers, one stack.

The honest answer

These are complements, and pretending otherwise would be marketing. Terra moves wearable data into your backend. Sahha turns signals into scores with contributing factors. CoachLayer turns training and recovery data into the sentence the athlete actually acts on: train, modify or rest, and why, with the numbers.

If you only need one layer, buy that layer. The comparison worth reading is what each layer refuses to do, because that is where teams lose quarters.

The stack

Data, scores, coaching: who stops where.

LayerVendorShipsDeliberately does not shipPublic pricing
Data pipeTerraNormalized health and fitness data from 500+ wearables and apps, delivered by API and webhooks, plus derived scores.Interpretation. The payload is normalized data and derived metrics, and what it means for training is your problem.From $399/mo on the annual plan ($499 monthly), 100,000 credits included, 200 credits per connected data source per user per month, overage from $0.005 per credit.
ScoresSahhaPhone and wearable signals turned into 0 to 100 scores (sleep, activity, readiness among them) with contributing factors, plus a trends and comparisons insights API.The narrative. A readiness of 62 with three contributing factors still needs someone to write the brief and the training decision.A 30-day free trial, then from $299/mo billed annually ($399 monthly) with 1,000 users included and $0.40 per user beyond.
CoachingCoachLayerReadiness briefs with an explicit train, modify or rest directive, workout analysis, weekly reports, adaptive programming, coach chat. Benchmarked on production traffic.Data acquisition. We do not connect to a single wearable; your data arrives in the request, from your backend or from a pipe like the ones on this page.Credit-metered tiers on the pricing table, sandbox free

Vendor facts checked October 2026 (Terra) and October 2026 (Sahha); list prices are theirs, not our restatement of a sales quote.

How it composes

Scores in, coaching out.

Terra + CoachLayer

Terra webhooks land sleep, HRV and activity in your backend; your backend passes them with the training log into a readiness brief call. Terra is the supplier lane here, not the rival one.

Sahha + CoachLayer

Sahha’s scores and factors make excellent structured input for the brief: the score says 62, the brief says what to do about it in the athlete’s language. Nothing in either contract fights the other.

Neither + CoachLayer

Plenty of apps already hold their training data and HealthKit reads. Then no pipe is needed at all: the coaching layer runs on what you have, and a thin data layer can come later.

What would change this page

The up-stack question, answered honestly.

The data and score vendors are visibly climbing toward interpretation: richer derived metrics, insight digests, trend narratives. As of September 2026, none of them ships narrative coaching with an explicit training directive as a product, which is the layer this page is about. If that changes, this comparison changes, and we will update it rather than defend it. Our bet is not that they cannot climb; it is that a coaching layer extracted from a shipped training product, with a production benchmark attached, is a different thing to catch up with than a feature checkbox.

Straight answers

Data pipe, scores or coaching: the questions buyers ask.

Do I need a wearable data API like Terra to use CoachLayer?

Only for data you do not already hold. CoachLayer never connects to a wearable: training and recovery data arrives in the request body, sent by your backend. If your app already logs workouts and reads Apple Health on the phone, you can call CoachLayer today with no pipe. If you need sleep, HRV and activity from many devices and vendor clouds, a pipe saves you one integration per source, and its output becomes CoachLayer’s input.

Can CoachLayer use Terra or Sahha output as input?

Yes, through your backend. Webhook payloads from a data pipe and scores from a scoring API land on your server; you map the fields you trust into the request, a sleep score and an HRV value into the readiness brief for example, next to the training log. CoachLayer never calls either vendor and holds no wearable credentials, so swapping the pipe later does not touch the coaching integration.

Can I compare HRV from Apple Health with WHOOP, Oura or Health Connect?

Not as raw numbers, because the platforms measure it differently. Apple’s HealthKit stores heart rate variability as SDNN, the standard deviation of the intervals between normal heartbeats, and Apple notes there are multiple ways of computing HRV. Health Connect’s HRV record on Android is RMSSD, the root mean square of successive differences between normal heartbeats. WHOOP’s API returns recovery HRV as RMSSD, and Oura says its app uses rMSSD, while its API describes the sleep value only as average heart rate variability during sleep. Apple, Health Connect and WHOOP all give the value in milliseconds, so nothing in the number tells you which method produced it. Keep each user on one source, store the metric with every value, and read HRV as a trend against that user’s own baseline. When a user switches device, start a new baseline on your side instead of sending the jump as a change in recovery. That trend, next to sleep and recent load, is what the readiness brief needs.

Is there a cloud API for Apple Health or Health Connect data?

No, and this is where wearable roadmaps slip. Neither Apple Health nor Health Connect offers a server API for a user’s records: an app on the phone reads them with the user’s permission and uploads them to your backend, which is why Terra routes both through its mobile SDKs. On Android the old cloud route is closing too: Google stopped new sign-ups to the Google Fit APIs on May 1, 2024, supports them only until the end of 2026, and says there is no alternative to the Fit REST API, pointing cloud integrations to the Google Health API instead. Budget for an on-device read either way.

Is wearable data health data under GDPR?

Treat it as such until your counsel says otherwise. GDPR Article 4(15) defines data concerning health as personal data related to the physical or mental health of a person which reveals information about their health status, and Article 9 prohibits processing it unless an exception applies, explicit consent for specified purposes being the first one listed. Resting heart rate, HRV or sleep tied to an identifiable user can qualify, so settle the lawful basis before you wire a pipe. On our side, request content is processed transiently: it is not written to a durable store and not used to train models.

Does HIPAA apply to wearable data in a fitness app?

Usually not, but US law still covers it. HIPAA applies to health plans, health care clearinghouses and health care providers that transmit health information electronically in covered transactions, plus the business associates handling data on their behalf. A consumer fitness app that works for none of them sits outside HIPAA and under the FTC’s Health Breach Notification Rule instead: its July 2024 amendments make clear that health apps and connected devices must comply, including apps that sync with a fitness tracker. A breach is not only a hack, since sharing the data without the user’s authorization also counts. Affected users must be notified within 60 calendar days, the FTC at the same time when 500 or more people are involved, and the FTC sets the civil penalty at up to $53,088 per violation. If you build for a clinic or an insurer and handle its patients’ data on its behalf, you are its business associate and HIPAA applies. On our side, request content is processed transiently: it is not written to a durable store and not used to train models.

How do data API prices compare with a coaching API?

On different units. Terra: From $399/mo on the annual plan ($499 monthly), 100,000 credits included, 200 credits per connected data source per user per month, overage from $0.005 per credit. Sahha: A 30-day free trial, then from $299/mo billed annually ($399 monthly) with 1,000 users included and $0.40 per user beyond. CoachLayer meters each coaching call in credits instead, 3 for a readiness brief, on paid tiers from $39 a month (Build), with 300 free Sandbox credits to evaluate. The pipe bill tracks connected users and the CoachLayer bill tracks the coaching calls you make, so budget them as two lines.

Can we connect wearables ourselves instead of paying for a data API?

Yes, one vendor at a time, and each vendor is its own project with its own terms. Garmin’s Health API is for businesses: you apply to the Garmin Connect Developer Program, get an evaluation environment after approval, and agree commercial terms Garmin tailors to each company. WHOOP’s API is currently free, but every developer on its platform needs a WHOOP device, which requires a membership. Oura’s API authenticates with OAuth2 and marks its personal access tokens as deprecated. Fitbit’s Web API was turned off on November 2, 2026 and now answers 410 Gone: integrations move to the Google Health API, whose scopes are all Restricted and need a privacy and security review before access. Each one brings its own sign-in flow, payload shape and approval timeline, and Apple Health still needs an on-device read. For one or two sources, going direct is reasonable; past that, a pipe such as Terra turns the list into one integration. Either way, the data lands on your backend and becomes CoachLayer’s input.

Vendor API terms: the Garmin Health API and Connect Developer Program pages, the WHOOP developer FAQ, Oura’s API authentication docs and Google’s Google Health API overview, checked November 2026.

US rules: 45 CFR 160.102 on who HIPAA applies to and the FTC’s Health Breach Notification Rule guide, checked November 2026.

HRV methods: Apple’s HealthKit SDNN reference, Health Connect’s RMSSD record, the WHOOP API recovery reference, Oura’s API docs and its note on rMSSD, checked November 2026.

Your data layer is solved. Coaching is the gap.

Whatever pipe you use, the brief is one POST away. Sandbox is free, no card.

Get a sandbox key